Failing O2 sensors

Status
Not open for further replies.

NJC

Joined
Apr 28, 2005
Messages
3,050
Location
Vancouver BC
Regarding my "Emission's failure - Volvo" post. I am considering testing my O2 sensor with an oscilloscope (if I can find) or a high impedance multimeter. Question: Wouldn't a failing O2 sensor give poor readings on both idle and driving tests? My tests show only idle failure (see top line):

Aircare2009.jpg


EDIT: O2 sensor is 10yrs old with 85K miles.
 
O2 sensors can be bad in a number of ways. They can slow down under certain condition, or give false readings, and still not throw a check engine light.
I believe you posted this problem elsewhere.
High HC is generally misfires, and high CO is generally rich.
A misfire can make it go rich. High CO generally does not make the HC go up much.
 
I did post the question elsewhere, but I'm trying to pinpoint specific modes of failure. For the cost ($~100), I may replace O2 sensor anyway.
 
Your O2 sensor may have a weak heater that may slow down response at idle.
You need to monitor min and max Voltage from your O2 as well as how fast it switches from low to high.
Lab scope it the way to go. Average DVOM probably isn't fast enough to monitor and they average the Voltage so you don't get accurate results.

At your age and mileage I would just replace it though. They don't last forever. 80K is abt average life span.
 
I think we do have oscilloscopes at work, but it doesn't look good having me walk out the door with it. I didn't realize they cycle so quickly (from link a few posts up):

Quote:
Most lambda sensors will cycle from rich to lean in about 50 to 100 milliseconds, and from lean to rich in 75 to 150 milliseconds. This is referred to as the “transition time”. If the O2 sensor is taking significantly longer to reverse readings, this too is an indication that it is getting sluggish and may need to be replaced.

The O2 sensor was replaced in 1999, or about 113k km's.
 
Originally Posted By: mechtech2

High HC is generally misfires, and high CO is generally rich.

Yes.

Originally Posted By: mechtech2

A misfire can make it go rich.


People argue this with me. A missfire causes a "lean" reading. All an O2 sensor does is measure the O2 content in the exhaust stream, the "rich" and "lean" is inferred. Missfire equals a lot of unburned fuel and O2 in the exhaust, and the excessive O2 means lean to the PCM so more fuel gets added to compensate. That overheats/kills a cat fast hence the requirement for a missfire monitor in OBDII.
 
punisher - A misfire allows the unburnt fuel to pass on to the O2 sensor.
Therefore it HAS to be rich.
This is why people argue with you.
 
Originally Posted By: mechtech2
punisher - A misfire allows the unburnt fuel to pass on to the O2 sensor.
Therefore it HAS to be rich.
This is why people argue with you.


The O2 sensor doesn't measure unburned fuel, it measures O2. Missfire=excess O2=rich.

Takes awhile. I argued with my instructer when he told it to me.
 
Originally Posted By: punisher
Missfire=excess O2=rich.

Quote:
Excessive HC results from ignition misfire or misfire due to excessively lean or rich air/fuel mixtures.

http://www.autoshop101.com/forms/h56.pdf

And excessive O2 would indicate a lean A/F mixture and would present a low voltage O2 sensor reading. When I pulled a few vacuum lines last night, I was able to achieve a solid reading of ~.2V.
 
Reading the excellent Picotech article on O2 sensors brought up one question.
Stoichometric mixture would mean complete combustion and, hence, no oxygen. How does a rich mixture cause even less oxygen in the exhaust?
My guess would be, in an imperfect world, even stoichmetric mixtures do not burn perfectly and some oxygen if left in the exhaust.
 
The typical O2 sensor is a binary device. It can only tell the computer whether the mixture is leaner than stoich or richer than stoich.

There are "wideband" O2 sensors that can tell the computer the exact mixture sensed, but these aren't that common. Most cars do not have them.
 
Originally Posted By: punisher
Missfire=excess O2=rich.


I correct it as pointed out above....excess 02=lean. Anytime you have excess O2 in the exhaust, no matter the source (missfire, actual lean condition, AIR injection valve failure) the PCM sees "lean" and will attempt to correct by adding fuel.
 
Last edited:
My 88 Mustang was giving lean codes due to a stuck thermactor air diverter solenoid. It was stuck diverting air upstream, before the O2 sensors.

The interesting thing is...I also needed to replace the fuel injectors because they were leaking externally. (Apparently a common problem with Bosch Design II injectors when they age).

The injectors came before the replacement thermactor air diverter solenoid. I replaced the injectors and re-ran the self test, expecting to get another lean code. Nope, this time I got a thermactor air stuck upstream code.

To this day I have no idea how it could tell the difference but only with a set of new injectors.
 
I borrowed a scope.
banana2.gif


fluke99.jpg

But can't figure out how to get any readings!
54.gif
frown.gif
Fiddle more on the weekend.
 
The scopemeter probaby has an auto mode where it will automatically select the sweep rate and volts per division to something that will give you an easy to read trace.

Try getting a reading off of something like a speaker or line output from a stereo, or a doorbell transformer or an AC adapter for an answering machine or cordless phone. Once you've figured out how to get a reading from one of those, then you can try getting a reading from your O2 sensor.
 
_2323_202chnlO2.gif


Here is a "typical" upstream (top) and downstream (bottom) O2 sensor wave form.
These are apparently from an Asian import.

You want to see Voltage switch from about 0.1-0.9v and switch quickly.

The setup on this scope is 0-1 volt and the time base is 1 second.
Put your Red probe on the sensor return wire and the Black on a good engine ground.
 
Brian, I couldn't get it to work on anything in the home ... but after some more experimenting I was able to get this.

img_3973.jpg


I don't know how to include the voltage units, but checking with a DMM rendered a cycling between ~.2V and ~.8V.
 
Status
Not open for further replies.
Back
Top